I'm seeking ideas on the sort of game mechanic that pits one character (or group of characters) against another character in a combat situation, but which doesn't involve fine-grained or low-latency changing of tactics. In other words, you have to prepare your character for the fight, and then send them in, waiting to see what happens. Basically this is because it's a text based game, so you have no on-screen positioning to play with, nor do you have real-time feedback on your actions. I'm hoping for something that rewards good choice of tactics before a fight on the part of the player, but which also allows for character (as opposed to player) advancement to affect the tactics available to you over time. Perhaps something like the Guild Wars system would work where you have a limited number of skills you can use at any one time, but can pick from a growing pool of them as you improve. In a combat setting I assume certain skills will work well against some opponents (and their own skills) and worse against others. It has to have more depth than simply trying to pick random or obscure skills and hoping for the best however, so any advice on achieving this would be welcome. What other 'preparatory' options are there? I'm hoping to allow players to choose different armour and weapon types based on predicting where and who they'll be fighting, with various rock-paper-scissors mechanics available for players who can make accurate predictions about their opponent. I also intend restricting what the character can carry so that they have to think more carefully before venturing forth, so there'll be no taking 7 different weapons and 3 different suits of armour for any possible occasion. Anything I've missed?

if you have a fairly large group of people it could be interesting to play it like a general, and decide on a battle plan but not to have to order every individual unit to move

for only a small party you could still have battle stratigy play a important part,

the only problem i see would be balancing, and trying to have a fairly large set of choice with without unbalanced ones that create a degenerate strategy

And whether it's an individual or small group tactics as mentioned above, this is going to be something where a user-customizable AI scripting will come in handy. The equipment is one thing, but tactics is something where you will be setting behavior threshold parameters. Much thought would have to go into creating an interface for those parameters and behaviors.

There may not necessarily be any complex tactics beyond the equipment and skill choice. I am certainly not envisaging much in the way of counterattack or instant response to attacks. Mainly I'm thinking about how to do it on a mathematical level, ie. how to set up a system that isn't purely down to your previous stats (eg. Dungeons and Dragons type systems), and on the other hand isn't purely down to real-time threat appraisal and response (eg. Soulcalibur/Tekken style games). Theoretically it could be decided instantly - as I was trying to imply with the title - so that it could be emitted as a single combat report, such as in a web-based game.

My game will feature control of a single character, sometimes fighting with allies, but don't let that stop anybody chiming in with more general ideas.

What about a phased based technique system with rules similar to tennis?
You have following phase and positions:

Initiate
Engagement
Finisher/Ace

The player assigns a technique to each of the above 4 slots. Each technique is used in a different phase and depending on the position that character is in.

Combat goes as follows:
Initiate techniques are used character who wins scores a point.
Engagement techniques are used until one character has at least 3 points and more points then the opponent.
Finisher phase then occurs the character with the most points then use their Finisher technique the other uses their Ace. If the Finisher wins the point then that character wins the battle otherwise combat moves back the engagement phase.

Combat can then be quick and easy to compute. Stamina could then be the main stat of interest as each technique could have a stamina cost and when stamina is depleted combat performance drops significantly.

Multiple opponents could be handled by having to defeat more then one opponent in each round, to score a win. Or you could add combos to enhance the advantages of fighting in groups.

The phases sound interesting, although it does seem to impose a requirement on there being an increased number of techniques, some phase-specific, while not necessarily increasing the range of strategies a player can use.

The strategy would come down to the choice of techniques, and what the different techniques did.

Initiate technique kick sand in eyes might apply -1 to all opponents success rolls if it hits but gives no points.

While the engagment technique grapple might have a low success rate but inflict the pinned status on your opponent and automatically moves to the finisher phase.

Quote:

Original post by Kylotan Perhaps something like the Guild Wars system would work where you have a limited number of skills you can use at any one time, but can pick from a growing pool of them as you improve.

I did a 2d RPG (if two screens and a boss count as a whole game) a while back with the similar mechanics in mind, but the player interface was standard (S)NES RPG fare.

The difference from Guild Wars was that I didn't use a fixed number of techniques, but I used a memory stat that determined how much you could remember in battle. Memory could be leveled up individually, but it cost much more than learning a new technique. Different techniques could cost different memory points, and there were bonuses for picking similar skills. I only implemented a couple of skill trees, but this fits the pattern you described.

As to rock-paper-scissors, I used a moderately complicated armor/penetration system. Slashing weapons do the most damage with the least penetration. Piercing weapons are medium in both categories, and blunt weapons do almost all of their damage regardless of armor (but cost the most in stamina). You could also throw in unique armors; i.e. chainmail offers similar protection to platemail with less weight for increased agility (good for avoiding blunt attacks), but is weak to piercing.

With those two above paragraphs in mind, a lot of the skills were used to convert one type of damage to another. For instance, "Thrust!" was only usable with swords, but it output piercing damage at 2/3 power instead of slashing at full power. Rapiers and jians actually do 1+1/3 damage, but this was all hard-coded, so I'm not sure how to implement this more robustly.

How are battles reported to the player? Is combat central to the game?

Quote:

Original post by templewulf The difference from Guild Wars was that I didn't use a fixed number of techniques, but I used a memory stat that determined how much you could remember in battle. Memory could be leveled up individually, but it cost much more than learning a new technique. Different techniques could cost different memory points, and there were bonuses for picking similar skills. I only implemented a couple of skill trees, but this fits the pattern you described.

Yes, that sounds a bit more elaborate than I was intending, but some aspects of that sound good. I may keep it simple in that you can typically only remember 8 skills in battle, but make it so that picking 3 from a certain classification allows you to remember a 4th for free, for example. So a benefit of this system is that picking your 8 memorised techniques is about more than just picking the best 8, but also having to choose between trading off 3 great skills for 4 skills that are less good on average. A negative is that I have to categorise all techniques, perhaps somewhat arbitrarily, while ensuring that each category is reasonably equal. Still, I can think of some ideas for the categories, some abstract - eg. the 4 elements, the 5 colours from Magic: The Gathering, 4 alignments from D+D - and some concrete - based on equipment type, based on location (city vs. wilderness), based on primary statistic (eg. Strength, Wisdom, Intelligence), etc.

Quote:

As to rock-paper-scissors, I used a moderately complicated armor/penetration system. Slashing weapons do the most damage with the least penetration. Piercing weapons are medium in both categories, and blunt weapons do almost all of their damage regardless of armor (but cost the most in stamina).

I hadn't thought of balancing the attack types in that way, but I was certainly considering having different armour types work better against the different piercing/slashing/bashing weapons. Obviously this only really adds gameplay value if you can predict or influence the sort of opposition you will come up against. Perhaps that could make stealth more important.

Quote:

How are battles reported to the player? Is combat central to the game?

In my game, battles will be reported in text form on a vaguely round by round basis, with the player able to use a limited number of options each time, such as communicating with allies, or attempting to flee. Combat is central to the game, but the emphasis is on the preparation for it rather than the act. Think of it as similar to soccer management, or training racehorses, or even something like Core Wars or robot soccer - you invest your time doing the best preparation you can, but your role during the actual resolution of the event is largely passive.

Quote:

Original post by Kylotan A negative is that I have to categorise all techniques, perhaps somewhat arbitrarily, while ensuring that each category is reasonably equal.

That was one of the reasons I ended up abandoning that design. I originally started with this classless system to get away from identical "builds", which was one of the things that drove me away from class-based RPGs, especially of the MMO variety. Giving bonuses to similar skills requires the designer to specify what qualities are similar, leading me right back to class types! I hate when my suggestions are obviated, especially by me! [grin]

Quote:

I hadn't thought of balancing the attack types in that way, but I was certainly considering having different armour types work better against the different piercing/slashing/bashing weapons. Obviously this only really adds gameplay value if you can predict or influence the sort of opposition you will come up against. Perhaps that could make stealth more important.

This system is typically better in SRPGs, where you have a party of adventurers, and the plate-armored knights can stand in front of the robe-wearing mages. In case the spearmen come out in force, you can switch your chainmail-wearing swordsmen to the second rank.

Typically games balance classes against each other, but you can't change classes with one adventurer (are comparisons to nethack appropriate?). Changing armor mid-mission doesn't sound like a good idea realistically, but Vagrant Story forced you to keep a small armory of weapons and shields suited to different opponents. Perhaps a little "research" with some classic games would help?

Quote:

Quote: How are battles reported to the player? Is combat central to the game? In my game, battles will be reported in text form on a vaguely round by round basis, with the player able to use a limited number of options each time, such as communicating with allies, or attempting to flee. Combat is central to the game, but the emphasis is on the preparation for it rather than the act. Think of it as similar to soccer management, or training racehorses, or even something like Core Wars or robot soccer - you invest your time doing the best preparation you can, but your role during the actual resolution of the event is largely passive.

This sounds a lot like the complaints I used to make about Dragon Quest and Final Fantasy. I think most of my time in FFIX was spent mashing the X button to get through inconsequential random battles, and I lamented how little tactical preparation counted. FFXII does the button mashing for you with gambits, so InnocuousFox's scriptable AI suggestion may be something to consider.